A method for linking classroom evaluations to specific physical properties
and for comparing the evaluations of different groups is described and illu
strated. Thirty-five college classrooms were photographed and shown to 20 p
rofessors and 51 undergraduate students, each of whom evaluated the friendl
iness of and their overall preference for all the classrooms. Seven physica
l properties of the classrooms were reliably assessed by independent observ
ers. Using a modified Brunswik lens model, the relations between the physic
al properties and the evaluations by the two groups were established and co
mpared. Between 40 and 57 per cent of the variance in the evaluations could
be explained from only three classroom properties: view to outdoors, seati
ng comfort and seating arrangement. Evaluations by the students and profess
ors were surprisingly similar, an encouraging sign for classroom designers.